Unit III: Fuzzy Logic: Concepts

and Foundations

Dr. S Ruban
Dept of Software Technology
 • Classical Relations:
• A classical relation is a subset of the Cartesian
product of two or more sets.
Introductio • Example: "is greater than" relation on a set of
numbers.
n to • These are crisp, where elements either belong to
Classical the relation or they do not.
• Fuzzy Relations:
and Fuzzy • Fuzzy relations allow degrees of membership.
Relations • Defined by a membership function with values
between 0 and 1.
• Example: "is tall" relation where height is evaluated
on a scale of 0 to 1.
Applications:
- Database systems for managing imprecise data.
- Image processing to define relationships between pixel
intensities.
Fuzzy Logic & Classical Logic
• Based on the principle of bivalence: A • Based on the principle of gradual
statement is either true or false. membership: A statement can be partially
true or partially false.
• Deals with absolute truth and falsity.
• Deals with degrees of truth and
• Uses precise values and clear-cut membership.
boundaries.
• Uses fuzzy sets and membership functions
• Limited in representing uncertainty to represent imprecise information.
and vagueness. • Better suited for modeling real-world
situations with uncertainty and ambiguity.
 • Definition:
Understand • The Cartesian product of two sets is the
set of all ordered pairs where the first element is
ing the from the first set and the second element is from
Cartesian the second set
• Overview
Product in • Cartesian product of two sets A and B
Relations • Forms the basis for creating relations.
• Example:
• A = {1, 2}, B = {x, y}
• Cartesian product: {(1, x), (1, y), (2, x), (2, y)}
Applications:
• Social networks to represent relationships (e.g.,
friendships).
• Product recommendation systems.
Cartesian
Product
 Reflexive

Symmetri
Transitive c
Tolerance Relations & Equivalence
Relations
• A binary relation that is reflexive and
• A binary relation that is reflexive,
symmetric. symmetric, and transitive.
• Represents a weaker form of • Represents a stronger form of equality or
similarity or compatibility between indistinguishability.
objects. • Example: "Two integers are considered
equivalent if they have the same
• Example: "Two people are considered remainder when divided by a given
similar if they share at least one number.“
common hobby."
• (2,1), (1,2) – Symmetric
• (1,2), (2,3), (3,1) - Transitive
• (1,1) - Reflexive
Property Tolerance Equivalence
Relation Relation
Reflexivity Yes Yes
Symmetry Yes Yes
Transitivity Not necessarily Yes
 • Tolerance Relation:
• Allows flexibility in defining similarity.
• Example: "Almost equal to" relation.
Tolerance • Used in approximate reasoning.
and • Equivalence Relation:
• Equivalence Relation is a type of relation that satisfies
Equivalenc three fundamental properties: reflexivity, symmetry,
and transitivity.
e Relations • Properties: Reflexive, Symmetric, Transitive.
• Example: Equality on a set of integers.
Applications:
• Clustering in machine learning.
• Categorization of objects based on attributes.
 Classical
Fuzzy Sets
Sets
• A classical set is a container that
• Fuzzy logic starts with the
wholly includes or wholly
concept of a fuzzy set. A fuzzy
excludes any given element.
set is a set without a crisp, clearly
defined boundary. It can contain • For example, the set of days of
elements with only a partial the week unquestionably
degree of membership. includes Monday, Thursday, and
Saturday. It just as
unquestionably excludes butter,
liberty, dorsal fins, and so on.
Features Fuzzy Set Crisp Set

Basic It is prescribed by vague or ambiguous properties. It is defined by precise and specific characteristics.

It is a set of components with different It is a set of objects that have the same
Definition
membership degrees in the set. countability and finiteness qualities.

Applications It is commonly utilized in fuzzy controllers. It is commonly utilized in digital design.

Membership It shows incomplete membership. It shows the complete membership.

Logic It follows the infinite-valued logic. It follows the bi-valued logic.

It specifies a number between 0 and 1, which

Value It specifies the value as either 0 or 1.
includes both 0 and 1.

Degree It defines the degree to which anything is true. It is also referred to as a classical set.
Membership
Function
• A membership function (MF) is a curve
that defines how each point in the
input space is mapped to a
membership value (or degree of
membership) between 0 and 1.
 • Definition:
• Membership function represents the degree to which
an element belongs to a fuzzy set.
Membershi • Values range from 0 to 1.

p Functions • Features:
• Shape: Triangular, Trapezoidal, Gaussian.
in Fuzzy • Flexibility to model real-world problems.
• Example:
Logic • For a fuzzy set "Young Age," membership value could be 1
for 20 years, 0.7 for 25 years, and 0 for 50 years.
Applications:
• Natural language processing to interpret vague terms.
• Control systems for adjusting inputs dynamically.
 Examples • Triangular Membership Function:
and • Simplicity in representation.
Application • Example: Age group as young, middle-aged,
old.
s of • Gaussian Membership Function:
Membershi • Smooth transition between degrees of
membership.
p Functions • Example: Speed categorization (slow,
medium, fast).
Applications:
• Image processing for edge detection.
• Decision-making in ambiguous situations.
 Features of Membership Functions

Core: The subset of Support: The range of

Boundary: The region
elements with full elements with any
where membership
membership in a fuzzy degree of membership
values transition.
set. in a fuzzy set.

Fuzzification: Defuzzification:
Converting crisp values Converting a fuzzy set
to fuzzy sets. back to a crisp value.
 • Definition:
• Process of converting crisp input into fuzzy values.
• Involves the use of membership functions.
• Methods:
• Linguistic approach: Converts input into terms like
Introduction "low," "medium," "high."
• Numerical scaling: Assigns membership values based
to on thresholds.
Fuzzification • Example:
• Temperature reading of 30°C could be fuzzified into
"warm" with a membership value of 0.8.
Applications:
• Weather prediction.
• Smart home devices for adaptive control.
 • Definition:
• Converts fuzzy output into a crisp value.
• Opposite of fuzzification.
• Common Methods:
Defuzzificat • Centroid of Area (COA).
• Mean of Maximum (MOM).
ion • Example:
Methods • In a fuzzy controller for an air conditioner, defuzzification
outputs a specific temperature setting.
Applications:
• Automotive systems like ABS.
• Robotics for motion control.
 Real-World
Use Cases • Industrial Automation:
• Example: Optimizing machinery operations.
of • Healthcare:

Defuzzificat • Example: Predicting patient recovery time based on

symptoms.
ion • Applications:
• Consumer electronics for adaptive performance.
• Traffic control systems for signal timing.
 • Definition:
Fuzzy • Arithmetic operations performed on fuzzy numbers.
• Includes addition, subtraction, multiplication, and
Arithmetic division.
Example:
and its Use • Adding two fuzzy numbers, e.g., "about 10" and "around
Cases 20" might yield "close to 30."
Applications:
• Financial modeling for uncertain forecasts.
• Risk analysis in project management.
Key points about fuzzy arithmetic

Fuzzy numbers: Fuzzy numbers are fuzzy sets with membership functions that peak
at a real number and assume positive values in an interval around that number.

Fuzzy sets: A fuzzy set is a class of objects with a continuum of grades of

membership.

Fuzzy variables: Fuzzy variables are variables whose states are fuzzy numbers.

Linguistic variables: When fuzzy numbers represent linguistic concepts, like "very
small", "small", or "medium", the final constructs are called linguistic variables.
 • Definition:
The • Generalizes classical functions to fuzzy functions.

Extension • Basis for fuzzy arithmetic and fuzzy control.

Example:
Principle in • Extending a function "y = 2x" to fuzzy inputs where x is
"about 5" to yield "about 10."
Fuzzy Logic Applications:
• Fuzzy control systems in engineering.
• Adaptive systems in AI.
 • Meaning
• "Fuzziness" in data refers to the presence of slight variations
or inaccuracies within data points, where information is not
perfectly precise and may include minor errors, misspellings,
Understand or inconsistencies, making it difficult to identify exact
matches when comparing different records, even if they
ing represent the same real-world entity
• Definition:
Fuzziness • Quantifies the level of uncertainty or ambiguity in a fuzzy
set.
in Data • Measures:
• Entropy-based measures.
• Distance-based measures.
Example:
• Measuring uncertainty in customer preferences.
Applications:
• Decision support systems.
• Market research for consumer behavior.
Introduction to Fuzzy Integrals
Definition: Fuzzy integrals are powerful aggregation operators that
can consider the interaction between criteria.
Example: Evaluating the importance of multiple criteria in
decision-making.
Applications: Environmental impact assessments. Multi-criteria
decision-making.
 • Definition:
• Decision-making under uncertainty using fuzzy logic.

Introductio • Incorporates subjective and vague data.

• Types:
n to Fuzzy • Individual Decision Making.
• Multiperson and Multiattribute Decision Making.
Decision • Example:
Making • Choosing the best investment portfolio considering "high
risk" and "moderate return."
• Applications:
• Supply chain optimization.
• Strategic business planning.
 Advanced • Multiobjective Decision Making:
• Example: Optimizing cost, quality, and time in
Fuzzy manufacturing.

Decision- • Multiattribute Decision Making:

• Example: Choosing the best product based on price,
Making quality, and durability.

Techniques • Applications:
• Urban planning for resource allocation.
• Business management for strategic decisions.
 • Definition:
Combinin • Integrates fuzzy logic with Bayesian
methods to handle uncertainty.
g Fuzzy
Logic with • Example:
Bayesian • Diagnosing diseases based on fuzzy
Probabilit symptoms and probabilistic evidence.
y • Applications:
• Healthcare diagnostics.
• Predictive modeling in finance.
 Fuzzy logic for vagueness:
Key points of • Fuzzy logic uses fuzzy sets, where elements can
Fuzzy logic belong to a set to a certain degree (not just fully in
or out), which is useful when dealing with
with imprecise or subjective information like "slightly
Bayesian hot" or "moderately likely."
probability Bayesian probability for updating beliefs:
• Bayesian probability allows for updating your
belief about a hypothesis based on new evidence,
utilizing Bayes' theorem to calculate posterior
probabilities.
 Fuzzy prior probabilities:
• Instead of crisp probabilities, you can use fuzzy
Working of membership functions to represent prior beliefs
about variables in a Bayesian model, reflecting
Fuzzy Logic uncertainty with linguistic terms.

with Fuzzy likelihood functions:

• When observing new data, the likelihood
Bayesian function can also be expressed as fuzzy
membership functions, allowing for a more
probability nuanced interpretation of how well the data
supports a hypothesis.
Fuzzy inference rules:
• Fuzzy logic rules can be incorporated into
Bayesian inference to reason with linguistic
information, combining the strengths of both
approaches.
 Practical Applications of Fuzzy Relations

Control Systems: Medical Diagnosis:

• Industrial process control: Adjusting parameters like temperature, • Symptom analysis: Interpreting a combination of symptoms with
pressure, or flow rate in industrial processes based on fuzzy rules varying degrees of relevance to diagnose diseases, particularly in
that incorporate "degrees of" concepts (e.g., "slightly too hot", areas where symptoms may overlap.
"moderately cold"). • Clinical decision support: Providing fuzzy-based recommendations to
• Vehicle control: Implementing fuzzy logic for advanced driving healthcare professionals based on patient data and medical
assistance systems, like adaptive cruise control or anti-skid braking, knowledge.
where situations are not always clear-cut.
• Home appliance control: Fuzzy logic in washing machines, air
conditioners, or vacuum cleaners to adapt operation based on
environmental conditions.

Image Processing: Decision Making Systems: Natural Language Processing:

• Image segmentation: Identifying boundaries • Expert systems: Incorporating fuzzy rules to • Semantic analysis: Interpreting the meaning
between objects in an image using fuzzy logic to represent expert knowledge and provide of words and phrases in a natural language
handle ambiguous pixel values. decision support in complex scenarios. context where degrees of truth can apply.
• Pattern recognition: Recognizing complex • Risk assessment: Evaluating risk levels in • Dialogue systems: Designing chatbots that
patterns in images by incorporating degrees of situations with uncertain information by can understand and respond to user queries
membership to handle variations. assigning fuzzy membership degrees to with nuanced meaning.
potential outcomes.
 Real-World Applications of Fuzzy
Decision Making
Home Appliances: Automotive Systems: Industrial Process Control:
• Washing machines: Adjusting wash cycles • Anti-lock braking systems: Optimizing • Chemical manufacturing: Maintaining
based on fuzzy inputs like "slightly dirty" or braking pressure based on road conditions optimal pH levels in a reaction by adjusting
"very dirty" by sensing the level of soiling. and vehicle speed. chemical inputs based on fuzzy logic
• Air conditioners: Automatically adjusting • Adaptive cruise control: Maintaining a safe • Power systems: Optimizing power
temperature based on "comfortable" or distance from the car ahead by considering distribution based on fluctuating demand
"cool" settings depending on ambient factors like traffic flow and road curvature.
conditions.

Medical Diagnosis: Robotics: Other Applications:

• Cancer detection: Analyzing medical • Obstacle avoidance: Enabling robots to • Financial trading: Making investment
imaging with fuzzy logic to identify potential navigate complex environments by decisions based on fuzzy analysis of market
cancerous areas considering "near" or "far" proximity to trends
• Anesthesia control: Adjusting anesthetic obstacles • Natural Language Processing: Interpreting
dosage based on patient vitals like heart • Path planning: Choosing the best route in a sentiment in text by assigning degrees of
rate and blood pressure dynamic environment based on fuzzy logic positivity or negativity to words
decision making
 Challenges and Limitations of Fuzzy
Logic
• 1. Ambiguity: Fuzzy Logic deals with imprecise and uncertain information, which can introduce ambiguity into
decision-making processes. This ambiguity arises due to the use of linguistic variables and fuzzy sets, which allow for a
more flexible representation of knowledge but can also make it challenging to precisely define boundaries.

• 2. Interpretation: Fuzzy Logic requires a careful interpretation of fuzzy rules and membership functions. The subjective
nature of linguistic variables and fuzzy sets can lead to different interpretations among decision-makers, potentially
impacting the consistency and reliability of decision outcomes.

• 3. Complexity: As the number of fuzzy rules and variables increases, the complexity of the decision-making process
grows. Managing a large number of rules and variables can become cumbersome and time-consuming, requiring
efficient algorithms and computational resources.

• 4. Data Acquisition: Fuzzy Logic heavily relies on data to define membership functions and establish relationships
between variables. Acquiring accurate and representative data can be a challenge, especially in dynamic business
environments where data availability and quality may vary.

• 5. Scalability: Scaling up fuzzy logic-based decision-making systems to handle large-scale problems can be a limitation.
The computational requirements and memory constraints associated with processing a vast amount of data and rules
can pose challenges in real-time decision-making scenarios.
Advancements and Innovations in
Fuzzy Logic
Type-3 Fuzzy Logic: Hybrid Fuzzy Systems: Fractional Calculus in Fuzzy Logic:
• This higher-order fuzzy logic system • Integrating fuzzy logic with other AI • Incorporating fractional calculus into
allows for more nuanced representation techniques like neural networks and fuzzy models enables better capture of
of uncertainty, making it suitable for genetic algorithms creates powerful long-range dependencies and non-local
complex decision-making scenarios in hybrid systems that can tackle complex interactions, leading to improved
robotics, power control, and classification problems with greater accuracy. prediction and control strategies for
tasks. complex systems.

Applications in Data Analytics and Adaptive Fuzzy Systems:

Image Processing: • Development of adaptive fuzzy systems
• Fuzzy logic is increasingly used in data that can learn and adjust their
analysis and image processing to enhance membership functions based on real-time
pattern recognition, feature extraction, data, improving system performance in
and decision-making processes. dynamic environments.
 • Integration with AI:
• Fuzzy logic combined with neural networks for
Advanceme better adaptability.
nts and • Applications in IoT:
Innovations • Smart home devices leveraging fuzzy systems
in Fuzzy for decision-making.
Logic
• Potential:
• Enhanced human-computer interaction.
• Improved decision-making in autonomous systems.
References
• https://www.mathworks.com/help/fuzzy/foundations-of-fuzzy-logic.h
tml

• https://www.javatpoint.com/difference-between-fuzzy-set-and-crisp-s
et

• https://www.slideshare.net/slideshow/a-presentation-on-fuzzy-equiv
alence-relations/124591071
Thank you